Elitebook 820 G3 engineering example: CMOS settings lost despite new cell

[i440bx]

Hey there,

I got a HP Elitebook 820 G3 Engineering Sample Laptop and I want to use it. Besides the difficult things at the ES-BIOS I have a big problem by saving the main BIOS settings.

The laptop does not have a big battery because it was dead, but I already plugged in a new CMOS battery.

It takes about 5min without power and the BIOS settings are lost. Unfortunatly I dont understand why the BIOS settings are only saved with power connected and this is the reason why I registered here.

As written above, I plugged in a new CMOS cell and checked the voltage (about 3.2V). During my research I found the chip the battery is connected to: nuvoton NPCE586HA0MX.

The chip has a 0 ohm connection to the coin cell connector, but it can't save the settings without power.



Do you have any idea where to look or can you point me in a direction where to look?

Thanks,
Robert

[mcplslg123]

RTCRST#/SRTCRST# present with only cmos cell connected,no power attached?

[i440bx]

RTCRST#/SRTCRST# present with only cmos cell connected,no power attached?

I read about these pins/conncetions but i did not found a Datasheet of the nuvoton chip.

Do you have this?

[piernov]

It's the southbridge that matters not the EC.

[i440bx]

It's the southbridge that matters not the EC.

Ah! So I need to find the pins on the southbridge and check if they are connected to the CMOS cell? This is the right hint, because I suspected the Nuvoton chip to be faulty.

Am I interpreting correctly that the CMOS setting is stored in the southbridge?

Edit:https://edc.intel.com/content/www/jp...s-and-signals/


Edit2: https://www.mouser.com/datasheet/2/6...fVYC_ZVs6_wPgq

[piernov]

Am I interpreting correctly that the CMOS setting is stored in the southbridge?

The actual BIOS settings are stored in the BIOS SPI ROM (a non-volatile memory), but the southbridge contains volatile registers (a long time ago some settings were actually stored there) that get cleared when the RTCRST# signal is triggered (i.e., goes to 0V) or the RTC power well is lost. The BIOS then detects a specific flag has been cleared and resets some of its settings during next boot.

[i440bx]

The actual BIOS settings are stored in the BIOS SPI ROM (a non-volatile memory), but the southbridge contains volatile registers (a long time ago some settings were actually stored there) that get cleared when the RTCRST# signal is triggered (i.e., goes to 0V) or the RTC power well is lost. The BIOS then detects a specific flag has been cleared and resets some of its settings during next boot.

Ahaaa! Thanks for explaining the process. So the problem is just to find the pins wich are probably pulled high or have a missing GND bridge.

As I dont have a shematics you maybe can hint me how to find these pins or the traces that are conected to? The Board has some solderjumpers but unfortnatly HP did only wrote the reference but not a name to them.


I know what is a SPI ROM At the ES Laptop its in a socket so I just need to find the right adapter/tool to read it.

[mcplslg123]

Tracing the RTC circuit with multimeter and without schematic needs experience.
There will be a diode(most likely 3 leg) near RTC cell. Check voltage on all legs with only cmos cell inserted.
However, its not necessary the diode will be close by.In that case you need to put your multimeter in continuity mode aka diode mode
and check continuity from RTC cell + side to where it beeps and further on.
Needs patience and experience. Good luck.

[i440bx]

Tracing the RTC circuit with multimeter and without schematic needs experience.
There will be a diode(most likely 3 leg) near RTC cell. Check voltage on all legs with only cmos cell inserted.
However, its not necessary the diode will be close by.In that case you need to put your multimeter in continuity mode aka diode mode
and check continuity from RTC cell + side to where it beeps and further on.
Needs patience and experience. Good luck.

I do have experience in electronics and troubleshooting, but nowhere near as complex circuits as this “new-fangled” mainboard.

Thank you very much for the tip, because it's enough to at least start tracing the wires.

Am I right in assuming that the button cell pulls up the RTCRST#/SRTCRST# pin, reduced in voltage by a diode? If so, one of the solder jumpers should also have the button cell voltage or the reduced button cell voltage applied.



€dit: Correction: If the voltage of the button cell were applied to the solder jumper (and the opposite side has GND) I would not have the problem of the forgotten CMOS settings.

I may have to find a closed solder jumper that fulfills the above requirements.

[mcplslg123]

RTCRST/SRTCRST# is part of RTC circuit inside PCH. yes the voltage is supllied by +RTCVCC.